Generative AI customer support assistant
Inspired by publicly reported work at Best Buy
Build an assistant that answers support questions, sorts requests and hands hard ones to a person.
The brief
Best Buy has written about using generative AI to improve customer support: helping customers troubleshoot, manage deliveries and subscriptions, and helping its agents find answers. Build your own support assistant for an electronics store: classify what the customer wants, answer from a knowledge base with sources, and hand over to a human with a summary when it should. Good looks like: intent accuracy on held-out messages, answers checked for grounding, a clear hand-over rule, and a test set of tricky and unsafe requests.
Read the case first
What Best Buy has said in public about this kind of work, on its own pages:
Make it your own
Build it for a different service desk: a college admissions office, a bike service centre or a telecom shop, in English plus one Indian language. Write the knowledge base yourself and decide which questions the assistant must never answer alone.
AITeky has no link with Best Buy, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as Best Buy's work or system.
What you will build
- Train an intent classifier on support messages
- Write a small knowledge base (returns, delivery, warranty, troubleshooting)
- Build retrieval-augmented answers that cite the knowledge base
- Add hand-over to a human with a conversation summary
- Evaluate intent accuracy, grounding and refusals on a test set
- Build a chat demo and an agent-side view
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.